leamsigc/shortsgenerator/facebook-research-template
Claude Code and agent skill by leamsigc/shortsgenerator
- community
Checking that the SKILL.md file is still in the repository.
Download the top 5 best-performing reels from a scraped Facebook page, transcribe them with Whisper, analyze the transcripts for structural patterns, and generate a reusable video template based on the best performer so the user can replicate or exceed its results.
Install
npx vskill@latest install leamsigc/shortsgenerator/facebook-research-templateSKILL.md
facebook-research-template — Video Template Extraction
Full pipeline: Download → Transcribe → Analyze → Template → Update REPORT.html
Prerequisites
openai-whisperinstalled (pip install openai-whisper)ffmpeginstalled- Chrome logged into Facebook (needed to access video pages)
- Previous states completed (scrape + analyze must have run)
Pipeline Steps
Step 1: Download Top 5 Reels
The script navigates to each of the top 5 posts (from page-analytics.json top5), extracts the <video> src, and downloads to video-template/downloads/post-N.mp4. If src is a blob URL or download fails, fall back to Chrome DevTools manual capture of the network request that serves the video.
Step 2: Transcribe with Whisper
Processes every file in video-template/downloads/ with Whisper base model (Spanish) and writes .txt + .srt to video-template/transcripts/. Skips already-transcribed files. Adjust --model in the script for speed (tiny) or accuracy (large).
Step 3: Analyze Transcripts
For each of the 5 transcripts, extract:
- Hook (0-5s): Exact wording, timing, tone
- Context (5-15s): How the situation is established
- Body/Drama (15-40s): The narrative arc, key moments
- CTA (40-50s): Call to action format
- Pacing: Sentence length distribution, questions per video, emphasis patterns
- Vocabulary: Power words, repeated phrases, emotional triggers
- Emotional arc: How the video makes the viewer feel moment-by-moment
Compare all 5 to find:
- What the best performer (#1) does differently
- Common patterns across all top performers
- What the worst performer (#5) is missing
Step 4: Generate Reusable Template
Create a fill-in-the-blank template based on the #1 post's structure.
The template includes:
- Exact hook formula with slots:
¡[STAR NAME] + [DRAMATIC VERB] + [TEAM]! [EMOJI] - Scene-by-scene structure with timing
- Vocabulary list (power words that work)
- CTA formula
- Production specs (duration, text style, audio)
Step 5: Update REPORT.html
Add a new section "🎬 Video Template" to REPORT.html showing:
- The reusable template
- The best performer breakdown
- Comparison table of all 5 transcripts
Output Files
| File | Description | |
|---|---|---|
video-template/downloads/post-1.mp4 through post-5.mp4 | Downloaded reels | |
video-template/transcripts/post-1.txt through post-5.txt | Whisper transcriptions | |
video-template/transcripts/post-1.srt through post-5.srt | Whisper subtitles with timestamps | |
video-template/transcript-analysis.json | Structured analysis of all 5 transcripts | |
video-template/VIDEO-TEMPLATE.md | Reusable fill-in-the-blank video template | |
video-template/best-performer-breakdown.md | Deep analysis of the winner |
Video Template Format
Hook: ¡[____] [____] [____]! [emoji] Context: [Team] estaba [situation]. Pero entonces... Drama: [Key moment] significó [impact] para [team]. Opinion: Lo que nadie dice es [controversial take]. CTA: ¿Crees que [debate question]? ¡Te leo! 👇
Notes
- If download fails (Facebook blocks), generate the template from the available metadata + analysis data
- Whisper model can be adjusted:
tiny(fast) tolarge(accurate) - For Spanish content,
basemodel works well; uselargefor accented speech
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewNot runA model review is scheduled for the next rescan.
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/facebook-research/facebook-research-template/SKILL.md
Popularity
- Repository stars
- 348
- Repository forks
- 75
- 7-day trend score
- 29
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/leamsigc/shortsgenerator/facebook-research-template)